zeeg contrasts mixed testing vs isolated prompt-only evals with OpenEval

zeeg · x · 2026-09-12

Continuing the exchange, zeeg notes OpenEval looks like a more constrained eval variant — isolated "test this prompt" runs — whereas his vitest-evals approach mixes evals into real testing workflows.

The thread highlights two design philosophies for agent evals: isolated agents with separate judge scoring vs. hybrid evaluation embedded in existing test suites.

Related event: Debating agent evals: traces over UI(2 posts)→

Original post →

More from coding & agent

coding & agent channel →